Predictive Analytics Framework for Traffic Congestion Using Artificial Intelligence (AI)Techniques

Abstract

Traffic congestion has continues to pose significant challenges in many urban centers globally, including Nigeria, undermining mobility, economic efficiency, and environmental sustainability. So far, conventional congestion management strategies in many developing countries is often focused on infrastructure expansion and manual traffic regulations, which have proven insufficient in addressing the complex, dynamic, and nonlinear behaviour of urban traffic systems. Therefore, this study examines traffic congestion control strategies across selected urban road networks in Nigeria, with a focus on operational traffic management, travel demand management, public transport prioritization, and developed an Intelligent Predictive Framework for Traffic Congestion using Artificial Intelligence to overcome these challenges. A multi-source and mixed method approach was employed; utilizing primary traffic data collected from major roads networks, signalized intersections during peak hours and historical traffic databases maintained by local transportation authorities. Congestion was assessed using indicators such as traffic volume, travel time, delay, speed, volume-to-capacity ratio, and level of service. The traffic dataset was partitioned chronologically into training (70%), validation (15%), and testing (15%) subsets. Experimental results demonstrated that the proposed AI-Based framework outperformed all other classifiers considered in this work achieving the lowest error values across all evaluation metrics with MAE (0.42), RMSE (0.58), and MAPE (6.9%), indicating superior predictive accuracy, robustness under peak-hour traffic conditions, and improved generalization to unseen data. Its performance gains are attributed to the hybrid integration of spatial and temporal learning, as well as optimized training and feature fusion strategies. Keywords: Intelligent Predictive Framework, Artificial Intelligence, Congestion Control, Intelligent Transportation Systems, Traffic Congestion. Aranuwa, F.O. (2026): Predictive Analytics Framework for Traffic Congestion Using Artificial Intelligence (AI)Techniques. Journal of Advances in Mathematical & Computational Science. Vol. 14, No. 1. Pp 13-30. Available online at www.isteams.net/mathematics-computationaljournal. dx.doi.org/10.22624/AIMS/MATHS/V14N1P2